Warmup that sends real conversations, not ping-pong

What warmup actually does to your reputation

Every large mailbox provider keeps a rolling reputation model for your sending domain and IP. It is fed by behavioural signals: how often recipients open your mail, how often they reply, how often they mark it as spam, how often they rescue it from spam, and how many of your sends bounce because the address never existed. None of these are things you can declare. They are things that have to happen.

Warmup produces those events deliberately. Warmerly enrols your mailbox in a network of other real mailboxes, then exchanges genuine mail with them on a schedule. The recipients open it, reply to it, and — critically — go and get it out of the junk folder when it lands there. Over a few weeks the provider's model shifts from 'unknown sender, treat with suspicion' to 'sender whose mail people read and answer'.

Why threaded conversations matter more than volume

The cheapest way to build a warmup network is to send a one-line message and have a bot reply with a one-line acknowledgement. It generates engagement events at very low cost. It is also trivially detectable: the messages have no quoted history, near-identical structure, uniform length, and reply latencies measured in seconds.

Warmerly's exchanges are built to look like the correspondence they are impersonating. A thread opens with a subject that reads like work, gets a reply that quotes it, gets an answer to the answer, and stops where a real thread would stop. Reply delays vary within human ranges. Message bodies differ between mailboxes rather than being drawn from one small pool. The point is not to trick anyone; it is that engagement earned by mail resembling your real mail is engagement that transfers to your real mail.

The junk-folder rescue loop

When a warmup message lands in the receiving mailbox's Spam or Junk folder, that is not a failure — it is the most useful moment in the whole cycle. Warmerly's processor connects to the receiving mailbox over IMAP, locates the message, moves it into the inbox, marks it as read, and sends the reply from the inbox thread.

To a provider, a human digging a message out of spam and answering it is close to the strongest positive signal available. It is a direct contradiction of the filter's own decision. A tool that only sends messages and never touches the receiving side cannot produce this signal at all, which is the practical difference between warmup that moves a domain and warmup that merely generates traffic.

How fast can you safely ramp?

Faster than most people fear, and much slower than most people attempt. The binding constraint is domain age. A domain registered last week has no history anywhere, and providers apply extra scepticism to young domains precisely because throwaway domains are the standard spam pattern. A domain that has been sending normal business mail for two years starts from a very different place.

Warmerly reads the real age of the domain and the mailbox and derives a daily allowance from it, then modulates that allowance by the mailbox's current health score and its measured spam placement. If placement degrades, the allowance narrows before you have done real damage. If everything is clean, it widens. You never write a ramp schedule, and there is no field where you can type a number that outruns your reputation.

Reputation-matched pairing

Not every mailbox in a warmup network is a good partner for every other mailbox. Pairing a brand-new domain with high volumes of traffic from mailboxes in poor standing spreads the problem rather than solving it. Warmerly pairs mailboxes by reputation band, so a healthy mailbox is not spending its budget corresponding with a domain that is currently listed, and a recovering mailbox is not thrown into the deep end on day one.

Because warmup exchanges are real mail between real mailboxes, the mailboxes on the other side can see your address, display name and signature — the same as any correspondent. That is stated plainly in the product rather than buried, because it is the honest trade for the network being made of genuine mailboxes instead of simulated ones.

What you see while it runs

The overview gives you a deliverability health score out of 100 averaged across your mailboxes, a rolling inbox-placement percentage split into inbox, spam and bounced, and today's sending against today's plan. Underneath, a per-weekday placement heatmap shows drift you would miss in a single number, and the mailbox table surfaces any account whose health has fallen far enough to need a look.

Each individual mailbox has its own page with the same figures scoped to it, plus its live SPF, DKIM, DMARC and MX status, its blocklist result, and a plain-English explanation of why today's limit is the number it is. The intent throughout is that every figure comes with the action it implies — a score with no next step is decoration.

Which mailboxes it works with

Google Workspace and Microsoft 365 connect over OAuth. Anything else connects with standard IMAP and SMTP credentials, which covers Zoho, Fastmail, Namecheap, IONOS, custom Postfix and effectively every business host. If you would rather not source and configure domains and mailboxes at all, Warmerly can host them for you and pre-configure the authentication records.

One caveat worth knowing before you start: Microsoft 365 tenants increasingly ship with SMTP AUTH disabled at the tenant level. OAuth will connect and IMAP will read, but sending fails with a 535 error until an administrator re-enables authenticated SMTP for the mailbox. That is a Microsoft setting, not a credential problem, and Warmerly tells you which of the two you are looking at rather than reporting a generic failure.